15) I do believe this is correct, this question is trying to trick you into thinking the magnets are super conductors! (If they were the magnetic field lines wouldn't pass through them!)
19) Yep! C is correct, A control is a very easy way to identify any uncontrolled variables.
20) This one is tricky! But in my opinion, I would go A. B doesn't really fit as placing no data because of a potential bias won't increase reliability. For an experiment, data is needed. Having peer-reviewed articles does increase reliability and does reduce this bias (and hence why I think A is better suited) as peer-reviewers are experts in the field, they know what they are talking about. Also, multiple peer-reviewers are present will increase this. Overall resulting in a more reliable report.
25) The general rule of thumb here is that the INDEPENDENT variable goes on x-axis, In this case, we aren't specifically told, but it can be safe to assume, that distance was the independent variable here
For part b, I will quickly draw up the graph and do b for you shortly